Condition monitoring method for rolling bearing, condition monitoring device, and program

Abstract

To more accurately capture signs of seizure in a rolling bearing and realize versatile condition monitoring.SOLUTION: A condition monitoring method for a rolling bearing comprises: an acquisition step of acquiring vibration data of the rolling bearing; an extraction step of extracting first data of a first time width from the vibration data; a generation step of generating second data of a second time width larger than the first time width by performing zero padding and amplitude correction on the first data; a derivation step of deriving a rolling-element passage frequency in the first data by performing frequency analysis on the second data; and a monitoring step of monitoring a condition of the rolling bearing by identifying signs of seizure occurring in the rolling bearing on the basis of a slope of change in the rolling-element passage frequency in the vibration data.SELECTED DRAWING: Figure 4

Description

[Technical Field]

[0001] The present invention relates to a rolling bearing condition monitoring method, a condition monitoring device, and a program. [Background technology]

[0002] Conventionally, rolling bearings have been used in various mechanical devices, and their condition is monitored to ensure proper operation and continuity of such devices. Examples of mechanical devices equipped with rolling bearings include wind power generation equipment and machine tools. One form of rolling bearing condition monitoring is monitoring for the occurrence of seizure. Seizure is often first detected when a rapid temperature rise occurs just before a serious failure of the rolling bearing. Therefore, there is a need for a method that enables detection of seizure at an earlier stage, rather than just before failure, in order to prevent failures and device malfunctions.

[0003] There are various methods for detecting seizure, such as methods using data on temperature and vibration magnitude, analysis of the deterioration state of lubricating oil, etc. For example, Patent Document 1 discloses a method for monitoring the condition of a rolling bearing by detecting a decrease in the rolling element passing frequency from vibration data of the rolling bearing. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2006-017291 Summary of the Invention [Problem to be solved by the invention]

[0005] On the other hand, conventional methods have issues such as the difficulty of predicting sudden burn-in, and although it is possible to detect signs of burn-in before it occurs (hereinafter referred to as burn-in precursors), it is expensive and difficult to deploy in actual equipment.

[0006] For example, wind turbines equipped with rolling bearings generally have slow rotational speeds, and therefore require extremely fine resolution to capture the frequency characteristics of the rotation. This requires long-term measurements of the operation of the wind turbine. When measurements are taken over a long period of time, the volume of measurement data becomes enormous, which inevitably increases the costs of calculation and storage.

[0007] Furthermore, a method is needed that can detect signs of general-purpose seizure regardless of the operating conditions and environment of the applied mechanical device. For example, when using a judgment method using a predetermined threshold for frequency as in Patent Document 1, the appropriate threshold depends on the actual operating conditions and environment of the mechanical device, making it difficult to set the threshold and, depending on the application environment, it may be difficult to accurately detect signs. Furthermore, measurement data is expected to contain noise, and the influence of such noise may make it impossible to extract the correct frequency characteristics (e.g., peak values) related to the operation of the rolling bearing. Therefore, simply comparing frequency characteristics based on measurement data affected by noise with thresholds as in Patent Document 1 may not fully capture physical phenomena that are signs of actual damage.

[0008] In view of the above problems, an object of the present invention is to more accurately detect signs of seizure in rolling bearings and to realize general-purpose condition monitoring. [Means for solving the problem]

[0009] In order to solve the above problems, the present invention has the following configuration: That is, a method for monitoring the condition of a rolling bearing, comprising: an acquisition step of acquiring vibration data of the rolling bearing; an extraction step of extracting first data of a first time width from the vibration data; a generating step of generating second data having a second time width greater than the first time width by performing zero padding and amplitude correction on the first data; a derivation step of deriving a rolling element passing frequency in the first data by performing frequency analysis on the second data; a monitoring step of monitoring a state of the rolling bearing by identifying a sign of seizure occurring in the rolling bearing based on a gradient of change in the rolling element passing frequency in the vibration data; A condition monitoring method comprising:

[0010] Another aspect of the present invention has the following configuration: A condition monitoring device for a rolling bearing, comprising: acquisition means for acquiring vibration data of the rolling bearing; extraction means for extracting first data of a first time width from the vibration data; generating means for generating second data having a second time width greater than the first time width by performing zero padding and amplitude correction on the first data; a derivation means for deriving a rolling element passing frequency in the first data by performing frequency analysis on the second data; a monitoring means for monitoring the state of the rolling bearing by identifying a sign of seizure occurring in the rolling bearing based on a gradient of change in the rolling element passing frequency in the vibration data; A condition monitoring method comprising:

[0011] Another aspect of the present invention has the following configuration: a program comprising: On the computer, an acquisition step of acquiring vibration data of the rolling bearing; an extraction step of extracting first data of a first time width from the vibration data; a generating step of generating second data having a second time width greater than the first time width by performing zero padding and amplitude correction on the first data; a derivation step of deriving a rolling element passing frequency in the first data by performing frequency analysis on the second data; a monitoring step of monitoring a state of the rolling bearing by identifying a sign of seizure occurring in the rolling bearing based on a gradient of change in the rolling element passing frequency in the vibration data; A program to execute. [Effects of the Invention]

[0012] The present invention makes it possible to more accurately detect signs of seizure in rolling bearings, thereby enabling general-purpose condition monitoring to be realized. [Brief explanation of the drawings]

[0013] [Figure 1] FIG. 1 is a block diagram showing an example of an apparatus configuration according to an embodiment of the present invention. [Figure 2] FIG. 10 is a graph illustrating a change in rolling element passing frequency when seizure occurs. [Figure 3] FIG. 10 is a graph illustrating zero padding. [Figure 4] 10 is a flowchart of a monitoring process according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0014] Hereinafter, embodiments for carrying out the present invention will be described with reference to the drawings. Note that the embodiment described below is one embodiment for explaining the present invention and is not intended to be interpreted as limiting the present invention. Furthermore, not all of the configurations described in each embodiment are necessarily essential configurations for solving the problems of the present invention. Furthermore, in each drawing, the same components are assigned the same reference numerals to indicate corresponding relationships.

[0015] Furthermore, in the explanations of this specification and the like, expressions such as "first" and "second" are used for convenience to distinguish from other elements, and are not intended to be interpreted as being limited to a specific configuration, etc. Therefore, the correspondence may be interpreted as appropriate depending on the components, etc. to which the present invention is applied.

[0016] First Embodiment A first embodiment of the present invention will be described below. The present invention can be used for condition monitoring of machinery equipped with rolling bearings, such as wind power generation equipment, machine tools, railways, and compressors. Therefore, the condition monitoring method according to the present invention is not particularly limited in terms of the type of equipment or rolling bearing to be monitored. In addition, types of rolling bearings to which the present invention can be applied include deep groove ball bearings, angular contact ball bearings, tapered roller bearings, cylindrical roller bearings, and self-aligning roller bearings.

[0017] Furthermore, the condition monitoring method according to this embodiment can be widely applied to machine tools that rotate at relatively high speeds and wind power generation equipment that rotate at relatively low speeds. In other words, the condition monitoring method according to this embodiment can be generally applied regardless of the rotational operation of the machine to which it is applied. As an example, conventional machine tools that rotate at high speeds use insulating bearings, such as ceramic ball bearings that do not conduct electricity, but the condition monitoring method according to this embodiment can be applied regardless of the electrical characteristics of such components. Note that the sampling period and measurement time of the measurement data are adjusted depending on the machine to which it is applied, as will be described in detail below.

[0018] [Device configuration] Fig. 1 is a schematic diagram showing an example of the overall configuration of an apparatus according to this embodiment. Fig. 1 shows the configuration of a bearing unit 100 to which a condition monitoring method according to this embodiment is applied, and a condition monitoring apparatus 200 that executes the condition monitoring method. For the sake of simplicity, Fig. 1 shows a configuration in which one bearing unit 100 is provided with one rolling bearing 101, but one bearing unit 100 may be provided with multiple rolling bearings 101. Furthermore, the bearing unit 100 may be provided with parts other than the rolling bearing 101, and for the sake of simplicity, only the configuration according to this embodiment is shown here.

[0019] The rolling bearing 101 rotatably supports the end of a rotating shaft (not shown). The rotating shaft is supported by a housing (not shown) that covers the outside of the rotating shaft via the rolling bearing 101, which is a rotating part. The rolling bearing 101 includes an inner ring 104, which is a rotating ring fitted onto the outside of the rotating shaft, an outer ring 102, which is a fixed ring fitted into the housing, a plurality of balls (rollers), which are a plurality of rolling elements 103 arranged between the inner ring 104 and the outer ring 102, and a cage 105 that holds the rolling elements 103 so that they can roll. Here, a configuration in which the outer ring 102 is fixed will be described as an example, but a configuration in which the inner ring 104 is fixed may also be used. The guide system of the cage 105 is not particularly limited, and may be any of an outer ring guide, an inner ring guide, or a rolling element guide.

[0020] Furthermore, in the rolling bearing 101, a predetermined lubrication method is used to reduce friction between the inner ring 104 and the rolling elements 103, and between the outer ring 102 and the rolling elements 103. The lubrication method is not particularly limited, but grease lubrication, oil lubrication, etc. are used, for example. The type of lubricant is also not particularly limited.

[0021] A vibration sensor 106 is installed in the housing of the bearing unit 100 so as to be able to detect vibrations from the rolling bearing 101 transmitted through the housing. The vibration sensor 106 detects vibrations from the housing and outputs them as vibration information to the condition monitoring device 200. The vibration sensor 106 may be any device capable of detecting vibrations, such as a microphone, acceleration sensor, AE (Acoustic Emission) sensor, ultrasonic sensor, or shock pulse sensor, as long as it can convert detected vibrations, such as sound, acceleration, velocity, strain, stress, or displacement, into an electrical signal. Furthermore, when installing the vibration sensor in a device located in a noisy environment, it is preferable to use an insulated type sensor, as it is less susceptible to noise. Furthermore, if the vibration sensor 106 uses a vibration detection element such as a piezoelectric element, the element may be molded in plastic or the like.

[0022] The condition monitoring device 200 may be configured, for example, by an information processing device such as a PC (Personal Computer). The condition monitoring device 200 includes an IF (Interface) unit 201, a processing unit 202, a storage unit 203, a UI (User Interface) unit 204, and a communication unit 205. The IF unit 201 is connected to the vibration sensor 106 and functions as an acquisition unit that acquires vibration information detected by the vibration sensor 106. The processing unit 202 may be configured, for example, by a CPU (Central Processing Unit), an MPU (Micro Processing Unit), a DSP (Digital Single Processor), or a dedicated circuit. The storage unit 203 is configured by volatile and non-volatile storage media such as a HDD (Hard Disk Drive), a ROM (Read Only Memory), and a RAM (Random Access Memory), and is capable of inputting and outputting various information in response to instructions from the processing unit 202. The processing unit 202 may execute the condition monitoring process according to this embodiment by, for example, reading and executing programs and various data stored in the storage unit 203.

[0023] The UI unit 204 is composed of a speaker, a light, a display device such as a liquid crystal display, and the like, and notifies the user in response to instructions from the processing unit 202. The notification method used by the UI unit 204 is not particularly limited, and may be, for example, an auditory notification using voice or a visual notification using screen output. The communication unit 205 is a network interface equipped with a communication function, and performs communication processing by transmitting data to an external device (not shown) via a network (not shown).

[0024] The condition monitoring device 200 may be provided integrally with a control device (not shown) that controls the rotational operation of the bearing unit 100, or the control device and the condition monitoring device 200 may be configured as separate devices. Here, an example is shown in which the control device and the condition monitoring device 200 are configured as separate devices.

[0025] [Changes in frequency related to image retention] Fig. 2 is a graph illustrating the change in frequency, particularly the rolling element passing frequency (Zfc), associated with the occurrence of seizure in the rolling bearing used in this embodiment. In Fig. 2, the vertical axis represents frequency [Hz], and the horizontal axis represents time [h]. The following description focuses on the case where the outer ring 102 constituting the rolling bearing 101 is fixed, and only the inner ring 104 rotates. In this case, Zfc of the outer ring 102 can be derived using the following formula (1). The constants in the following formula are predetermined according to the specifications of the rolling bearing 101.

[0026]

number

[0027] In Figure 2, if the time when rotation started is set to 0, seizure occurs in the rolling bearing 101 after about 70 hours. After seizure occurs, rotation is stopped. Here, it can be seen that just before seizure occurs and the operation of the rolling bearing 101 stops (range P2), Zfc decreases at an accelerated rate, that is, suddenly. If such a sudden decrease in Zfc can be detected, it can be treated as a sign that seizure is about to occur.

[0028] On the other hand, if we focus on the range P1 from around 30 hours to 65 hours in Figure 2, we can see that the Zfc value gradually decreases, unlike the sudden decrease in the range P2. If we can capture this kind of decrease, we can detect signs of burn-in earlier. In order to detect this gradual decrease in Zfc, it is necessary to increase the frequency resolution so that even slight changes in Zfc can be captured. As mentioned above, when increasing the resolution, the measurement time for the measurement data is generally extended. However, simply extending the measurement time leads to issues such as increased calculation costs and storage costs, and the inability to capture sudden changes such as those in the range P2.

[0029] For example, in the case of a spindle bearing of a machine tool that rotates at high speed, Zfc can reach a high value of 3 kHz or more during high-speed rotation, partly due to the large number of balls. Under high-speed rotation conditions, not only is vibration noise large, but the rolling elements pass by 3,000 times or more in a short period (for example, 1 second) around the measurement position of the vibration data, which can cause variations in the signal value. As a result, the peak indicating Zfc becomes less noticeable in the frequency analysis of the vibration data.

[0030] Furthermore, in the case of spindle bearings of machine tools that rotate at high speeds, seizure is likely to occur soon after signs of seizure are observed due to the high rotation speed. In other words, the faster the rotation speed, the shorter the interval between ranges P2 shown in Figure 2 tends to be. For this reason, it is necessary to reliably capture sudden changes in Zfc such as those in range P2, while also quickly and accurately capturing the signs of the preceding stages in range P1.

[0031] The method according to this embodiment anticipates the following changes in the state of the monitored object and detects signs of seizure. When a monitored object uses a lubricant, poor lubrication occurs due to lubricant deterioration (e.g., oxidation), a lack of lubricant, or the presence of foreign matter. When poor lubrication occurs, the oil film breaks between the components of the monitored object. Examples of between the components include the rolling element and the raceway, and the rolling element and the cage pocket.

[0032] When the oil film is broken, contact occurs between the components, resulting in minute friction and wear. Friction and wear between the components increases the amount of tiny debris (wear particles) in the lubricant, causing adhesive wear between the components. As wear increases, the passing frequency of the rolling elements decreases. By detecting the frequency at this time, i.e., the fluctuation (decrease) in Zfc, it is possible to detect signs of seizure.

[0033] In this embodiment, a zero-padding FFT (Fast Fourier Transform) is used as a method for capturing finer changes in Zfc.

[0034] [Zero-padded FFT] 3 is a graph illustrating an application example of the zero-padding FFT used in this embodiment. Since the zero-padding FFT technique is well known, it will be briefly explained here. In the zero-padding FFT, in order to improve frequency resolution in short-time measurements, for example, zero-padding processing and amplitude correction processing are applied to vibration data detected by the vibration sensor 106, and then FFT processing is applied.

[0035] In FIG. 3(a), the vertical axis represents amplitude and the horizontal axis represents time. As an example, waveform 301 is vibration data detected by vibration sensor 106 over a short period of time. In order to obtain accurate amplitude components, zero padding is performed on this, as shown in waveforms 302 and 303. This converts the signal value into a waveform with a sufficient time length. In other words, the apparent frame period is lengthened. If an FFT is performed on the signal value in this state, the frame period is simply lengthened, and the amplitude relative to the frequency is obtained as a small value. In response to this, the signal value is further corrected using an amplitude correction coefficient.

[0036] FIG. 3(b) shows a graph of the results of amplitude correction processing performed on the signal values shown in FIG. 3(a). In FIG. 3(b), the vertical axis indicates amplitude and the horizontal axis indicates time, but the scale of the vertical axis is different. In other words, the amplitude of the signal values shown in FIG. 3(a) is increased by the amplitude correction coefficient. Note that the signal values in the range added by zero padding, i.e., the amplitude values, remain 0.

[0037] FIG. 3(c) is a graph showing the results of FFT processing. The vertical axis represents amplitude, and the horizontal axis represents frequency [Hz]. Waveform 321 shows the results when FFT is applied to signal values obtained over a time width sufficient to obtain frequency resolution. Peaks can be identified in waveform 321. In this embodiment, to accurately capture changes in Zfc, peaks are detected based on signal values over a short period of time, instead of vibration data with a sufficient time width such as that obtained in waveform 321. In this case, in this embodiment, a signal with a desired time width is derived in a simulated manner using the zero padding described above. This makes it possible to detect peaks such as those in waveform 321, even with signal data with a short time width.

[0038] Waveform 322 shows the result of applying FFT to the signal value acquired by vibration sensor 106. In this case, the time width of the signal is not sufficient for the signal waveform, so accurate amplitude components cannot be acquired. In other words, the appropriate peak cannot be identified.

[0039] Waveform 323 shows the result obtained by applying an FFT to waveform 311 shown in FIG. 3(b), which was obtained using the zero padding and amplitude correction coefficients described above. Waveform 323 allows peak detection, just like waveform 321. In other words, performing amplitude correction as shown in FIG. 3(b) makes it possible to achieve accurate FFT intensity calculations. Furthermore, since peak detection is possible at shorter time intervals, it is possible to improve frequency resolution.

[0040] [Status monitoring process] 4 is a flowchart of the condition monitoring process according to this embodiment. This process is executed by the condition monitoring device 200, and may be realized, for example, by the processing unit 202 included in the condition monitoring device 200 reading out from a storage device and executing a program for implementing this process. For ease of explanation, the processing will be collectively described as being performed by the condition monitoring device 200. This process may be started when the rolling bearing 101 starts to rotate, or may be started based on a user instruction. In either case, it is assumed that the rolling bearing 101 is rotating when this process flow is executed.

[0041] In S401, the state monitoring device 200 acquires, as vibration data, a signal detected by the vibration sensor 106. In this acquisition process, A / D conversion, noise removal, envelope processing, etc. may be further performed.

[0042] In S402, the condition monitoring device 200 extracts a signal of a predetermined time width from the vibration data acquired in S401. This predetermined time width is assumed to be defined in advance and will hereinafter also be referred to as a "first time width." In the example of FIG. 3, a time width corresponding to waveform 301 is defined.

[0043] In S403, the state monitoring device 200 performs zero padding on the vibration data extracted in S402. Specifically, as shown in Fig. 3(a), the state monitoring device 200 performs zero padding (adding waveforms 302 and 303) on the acquired vibration data (waveform 301) to expand the time width. Here, the first time width is adjusted to become the second time width.

[0044] In S404, the condition monitoring device 200 performs frequency analysis by applying FFT processing to the vibration data obtained in S403. Then, the condition monitoring device 200 derives each parameter by identifying the peak value of the waveform from the obtained results.

[0045] In S405, the state monitoring device 200 calculates Zfc using the parameters obtained in S404 and the calculation formula defined by the above formula (1).

[0046] In S406, the state monitoring device 200 determines whether Zfc calculated in S405 has fallen below a predetermined threshold. The determination using the predetermined threshold in S406 is intended to detect a sign of burn-in just before it occurs. The predetermined threshold can be set to, for example, a value such as that shown as threshold Th in FIG. 2. The predetermined threshold is determined in advance and stored in the storage unit 203. If Zfc has fallen below the predetermined threshold (YES in S406), the processing of the state monitoring device 200 proceeds to S410. On the other hand, if Zfc is not below the predetermined threshold (NO in S406), the processing of the state monitoring device 200 proceeds to S407.

[0047] In S407, the condition monitoring device 200 calculates the average slope of Zfc over the first period using the least squares method. Since the least squares method is a well-known technique, detailed description thereof will be omitted here. The first period is a predetermined unit time of vibration data, within which multiple Zfc are derived. For example, the first period is 30 minutes. The average slope of the multiple Zfc calculated over the first period is then calculated using the least squares method. More specifically, if Zfc is calculated every minute over a 30-minute period, an average slope based on 30 Zfc is calculated. Furthermore, if Zfc is calculated every 5 minutes over a 30-minute period, an average slope based on 6 Zfc is calculated. The number of times (cycle) calculated over the first period may correspond to the first time width of S402.

[0048] In S408, the state monitoring device 200 determines whether the slope of Zfc calculated in S407 satisfies a predetermined condition. The predetermined condition here may be defined as follows, for example. (Condition 1) If the average slope of Zfc in the first period is negative for the second period consecutively, it is determined to be a sign of high probability of burn-in (second period > first period). (Condition 2) If the average slope of Zfc in the first period remains negative for the third period, it is determined to be a sign of a medium possibility of image sticking (second period > third period > first period).

[0049] Specifically, as an example, it can be defined as follows: (Example of condition 1) If the first period is 30 minutes, the second period is 60 minutes, and Zfc is calculated every minute, 30 Zfc (= 30 minutes / 1 minute) are used to calculate the average slope of Zfc every minute, and if this average slope is negative 60 times in a row (= 60 minutes / 1 minute), it is detected as a sign of a high possibility of burn-in. (Example of condition 2) If the first period is 30 minutes, the third period is 40 minutes, and Zfc is calculated every 5 minutes, six Zfc (= 30 minutes / 5 minutes) are used to calculate the average slope of Zfc every 5 minutes, and if this average slope is negative eight times in a row (= 40 minutes / 5 minutes), the possibility of burn-in is detected as a moderate sign.

[0050] From the viewpoint of improving accuracy, it is preferable that the number of Zfc for calculating the average slope of Zfc is six or more. In other words, it is preferable to calculate Zfc by taking measurements more frequently than once every five minutes. Furthermore, under high-speed rotation conditions, for example, it is preferable that a single measurement lasts for 0.5 to 5 seconds with a sampling frequency of 10 kHz or more. The number of seconds here corresponds to the first time width in S402. Furthermore, considering low-speed rotation conditions, it is preferable that a single measurement lasts for X to 100X seconds with a sampling period of Y kHz or more (X=100 / theoretical Zfc, Y=5×theoretical Zfc).

[0051] In addition, although two conditions are given as examples in the above example, more conditions may be defined. Also, for example, by setting a plurality of different parameters based on condition 1, conditions may be set so that the degree (possibility) of signs of image sticking can be determined in stages.

[0052] In S508, the state monitoring device 200 determines whether the calculated slope of Zfc satisfies the above-mentioned predictive conditions. If the slope of Zfc satisfies any of the conditions (YES in S408), the processing of the state monitoring device 200 proceeds to S410. On the other hand, if the slope of Zfc does not satisfy the conditions (NO in S408), the processing of the state monitoring device 200 proceeds to S409.

[0053] In S409, the state monitoring device 200 stores the Zfc calculated in S405 in the storage unit 203. The Zfc stored here is used for the subsequent calculation in S407. Note that the value of Zfc may be periodically deleted from the storage unit 203 after a certain period of time has passed since the calculation. Then, the processing of the state monitoring device 200 returns to S401, and state monitoring continues.

[0054] In S410, the state monitoring device 200 detects a sign of burn-in. At this time, the state monitoring device 200 may also determine whether the determination using the first threshold in S406 detected a sign immediately before the occurrence of burn-in, or whether the determination using the average slope of Zfc in S408 detected a sign a certain period before the occurrence of burn-in. Then, the processing of the state monitoring device 200 proceeds to S411.

[0055] In S411, the condition monitoring device 200 notifies information related to the detected signs of seizure. The condition monitoring device 200 may notify that a sign of imminent seizure has been detected based on the determination result of S406. The condition monitoring device 200 may also notify in stages that the possibility of seizure occurrence has been detected based on the determination result of S408. The notification method here is not particularly limited. For example, the user may be notified visually or audibly. Alternatively, a control device (not shown) of the rolling bearing 101 may be notified to stop operation of the rolling bearing 101. Then, this processing flow ends.

[0056] In the above processing flow, if a sign of seizure is detected, a notification to that effect is sent and the processing is terminated. However, this is not limited to this, and even if a sign of seizure is detected and a notification is sent, the condition monitoring may continue as long as the rolling bearing continues to operate.

[0057] As described above, this embodiment makes it possible to more accurately detect signs of seizure in rolling bearings and realize general-purpose condition monitoring. In particular, by detecting changes in Zfc, it is possible to accurately monitor the internal condition of the rolling bearing. For example, it becomes possible to monitor the process leading up to the occurrence of seizure, which cannot be detected by the temporary magnitude of temperature or vibration alone. Furthermore, by using a zero-padding FFT, it is possible to detect changes in Zfc with high resolution while suppressing the volume of vibration data required for condition monitoring.

[0058] <Other embodiments> In the present invention, a program or application for realizing the functions of one or more of the above-described embodiments can be supplied to a system or device using a network or a storage medium, etc., and one or more processors in the computer of the system or device can read and execute the program.

[0059] As such, the present invention is not limited to the above-described embodiments, and the present invention also contemplates the mutual combination of the various components of the embodiments, as well as modifications and applications by those skilled in the art based on the description in the specification and well-known techniques, and these modifications and applications are included in the scope of protection sought.

[0060] As described above, the present specification discloses the following: (1) A method for monitoring the condition of a rolling bearing (e.g., 101), comprising: an acquisition step of acquiring vibration data of the rolling bearing; an extraction step of extracting first data of a first time width from the vibration data; a generating step of generating second data having a second time width greater than the first time width by performing zero padding and amplitude correction on the first data; a derivation step of deriving a rolling element passing frequency in the first data by performing frequency analysis on the second data; a monitoring step of monitoring a state of the rolling bearing by identifying a sign of seizure occurring in the rolling bearing based on a gradient of change in the rolling element passing frequency in the vibration data; A condition monitoring method comprising: This configuration allows for more accurate detection of signs of bearing seizure, enabling general-purpose condition monitoring. In particular, by detecting changes in the rolling element pass frequency, the internal condition of the rolling bearing can be monitored with high accuracy. Furthermore, by using zero-padding FFT, it is possible to detect changes in the rolling element pass frequency with high resolution while reducing the volume of vibration data required for condition monitoring.

[0061] (2) The condition monitoring method according to (1), wherein the monitoring step detects a case where the gradient of the change in the rolling element passing frequency per unit time in the vibration data is a negative value continuously for a predetermined period of time as a sign of seizure. According to this configuration, by detecting signs of rolling bearing seizure based on the negative continuity of the slope of the change in rolling element passing frequency per unit time, it becomes possible to monitor the condition of the rolling bearing independently of the operating environment or operating conditions.

[0062] (3) The condition monitoring method according to (2), wherein the monitoring step uses the least squares method to calculate an average slope of the rolling element passing frequency per unit time from the values of a plurality of rolling element passing frequencies derived for each predetermined period using the vibration data for the unit time. According to this configuration, the gradient of the change in the rolling element passing frequency per unit time is derived using the least squares method, making it possible to perform condition monitoring through simple processing.

[0063] (4) A condition monitoring method according to any one of (1) to (3), wherein the monitoring step further detects that a rolling element passing frequency derived based on the vibration data falls below a predetermined threshold value as a sign that seizure is likely to occur. According to this configuration, it is possible to detect a sudden drop in the rolling element passing frequency as a sign immediately before the occurrence of seizure in the rolling bearing.

[0064] (5) The condition monitoring method according to any one of (1) to (4), wherein the rolling element passing frequency is a peak frequency of Zfc, which is a frequency component of an outer ring constituting the rolling bearing. According to this configuration, it is possible to detect the signs of image sticking by focusing on Zfc.

[0065] (6) A condition monitoring method according to any one of (1) to (5), wherein the rolling bearing is made of an insulating material. With this configuration, it is possible to perform condition monitoring even for rolling bearings that have characteristics to which a condition monitoring method using oil film measurement results by applying a current, such as the electrical impedance method, cannot be applied.

[0066] (7) A condition monitoring device (e.g., 200) for a rolling bearing (e.g., 101), Acquisition means (e.g., 106, 201) for acquiring vibration data of the rolling bearing; An extraction means (e.g., 202) for extracting first data of a first time width from the vibration data; A generating means (e.g., 202) for generating second data having a second time width greater than the first time width by performing zero padding and amplitude correction on the first data; a derivation means (e.g., 202) for deriving a rolling element passing frequency in the first data by performing frequency analysis on the second data; a monitoring means (e.g., 202) for monitoring the state of the rolling bearing by identifying a sign of seizure occurring in the rolling bearing based on the gradient of change in the rolling element passing frequency in the vibration data; A condition monitoring device having: This configuration allows for more accurate detection of signs of bearing seizure, enabling general-purpose condition monitoring. In particular, by detecting changes in the rolling element pass frequency, the internal condition of the rolling bearing can be monitored with high accuracy. Furthermore, by using zero-padding FFT, it is possible to detect changes in the rolling element pass frequency with high resolution while reducing the volume of vibration data required for condition monitoring.

[0067] (8) To a computer (e.g., 200), an acquisition step of acquiring vibration data of a rolling bearing (e.g., 101); an extraction step of extracting first data of a first time width from the vibration data; a generating step of generating second data having a second time width greater than the first time width by performing zero padding and amplitude correction on the first data; a derivation step of deriving a rolling element passing frequency in the first data by performing frequency analysis on the second data; a monitoring step of monitoring a state of the rolling bearing by identifying a sign of seizure occurring in the rolling bearing based on a gradient of change in the rolling element passing frequency in the vibration data; A program to execute. This configuration allows for more accurate detection of signs of bearing seizure, enabling general-purpose condition monitoring. In particular, by detecting changes in the rolling element pass frequency, the internal condition of the rolling bearing can be monitored with high accuracy. Furthermore, by using zero-padding FFT, it is possible to detect changes in the rolling element pass frequency with high resolution while reducing the volume of vibration data required for condition monitoring. [Explanation of symbols]

[0068] 100...Bearing unit 101...Rolling bearing 102...Outer ring 103...Rolling element 104...Inner circle 105...Cage 106...Vibration sensor 200...Condition monitoring device 201...IF (Interface) section 202...Processing section 203...Storage section 204...UI (User Interface) section 205…Communications Department

Claims

1. A method for monitoring the condition of a rolling bearing, comprising: an acquisition step of acquiring vibration data of the rolling bearing; an extraction step of extracting first data of a first time width from the vibration data; a generating step of generating second data having a second time width greater than the first time width by performing zero padding and amplitude correction on the first data; a derivation step of deriving a rolling element passing frequency in the first data by performing frequency analysis on the second data; a monitoring step of monitoring a state of the rolling bearing by identifying a sign of seizure occurring in the rolling bearing based on a gradient of change in the rolling element passing frequency in the vibration data; A condition monitoring method comprising:

2. 2. The condition monitoring method according to claim 1, wherein the monitoring step detects a case in which a gradient of change in the rolling element passing frequency per unit time in the vibration data is a negative value continuously over a predetermined period as a sign of seizure.

3. 3. The condition monitoring method according to claim 2, wherein the monitoring step uses a least squares method to calculate an average slope of the rolling element passing frequency per unit time from values of a plurality of rolling element passing frequencies derived for each predetermined period using the vibration data for the unit time.

4. 2. The condition monitoring method according to claim 1, wherein the monitoring step further comprises detecting, when the rolling element passing frequency derived based on the vibration data falls below a predetermined threshold, that this is a sign that seizure is highly likely to occur.

5. 2. The condition monitoring method according to claim 1, wherein the rolling element passing frequency is a peak frequency of Zfc, which is a frequency component of an outer ring constituting the rolling bearing.

6. The condition monitoring method according to claim 1 , wherein the rolling bearing is made of an insulating material.

7. A condition monitoring device for a rolling bearing, comprising: acquisition means for acquiring vibration data of the rolling bearing; extraction means for extracting first data of a first time width from the vibration data; generating means for generating second data having a second time width greater than the first time width by performing zero padding and amplitude correction on the first data; a derivation means for deriving a rolling element passing frequency in the first data by performing frequency analysis on the second data; a monitoring means for monitoring the state of the rolling bearing by identifying a sign of seizure occurring in the rolling bearing based on a gradient of change in the rolling element passing frequency in the vibration data; A condition monitoring device having:

8. On the computer, an acquisition step of acquiring vibration data of the rolling bearing; an extraction step of extracting first data of a first time width from the vibration data; a generating step of generating second data having a second time width greater than the first time width by performing zero padding and amplitude correction on the first data; a derivation step of deriving a rolling element passing frequency in the first data by performing frequency analysis on the second data; a monitoring step of monitoring a state of the rolling bearing by identifying a sign of seizure occurring in the rolling bearing based on a gradient of change in the rolling element passing frequency in the vibration data; A program to execute.

Images